Release 3.0.0 of Ethyca's Astralis platform records every purpose based access control decision with its outcome, reason, inferred data use and confidence, the tables and columns touched, and the policies that denied it. The records are exposed through new decision record API endpoints and a decision drawer in the enforcement log. Behind a flag that is off by default, the MCP gateway can also infer the data use of an AI agent's database query and decide on it.
Buyer relevance. Privacy and legal teams get a per decision audit trail showing why an AI agent or a user was allowed or denied access to sensitive data, which is the evidence an auditor or regulator will ask for.